Total Variation-Based Image Restoration Using I-Divergence
Qiufeng Wu, Kuanquan Wang, Wangmeng Zuo · International Review on Computers and Software (IRECOS) · 2013
This paper proposes I-divergence and TV-based image restoration model for recovering the sharp image from the blurred noisy image. This model is first formalized as the optimization problem, and then it is converted into Euler-Lagrange equation, which is solved by the gradient descent flow. The experiments verify that PSNR and BSNR of our I-divergence and TV based image restoration model is superior to those of the least squared criterion and TV based image restoration model in the identical problem.